[PATCH next] drm/sysfb: Fix a NULL vs IS_ERR() bug

[PATCH next] drm/sysfb: Fix a NULL vs IS_ERR() bug

[Dan Carpenter]

The drm_atomic_get_crtc_state() function never returns NULL, it returns
error pointers.  Update the error checking to match.

Fixes: cb71de092553 ("drm/sysfb: Lookup blit function during atomic check")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
---
 drivers/gpu/drm/sysfb/drm_sysfb_modeset.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
index 8517c490e815..d2c4d8f3d4d0 100644
--- a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
+++ b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
@@ -259,7 +259,7 @@ int drm_sysfb_plane_helper_begin_fb_access(struct drm_plane *plane,
 	ret = -EINVAL;
 
 	crtc_state = drm_atomic_get_crtc_state(plane_state->state, plane_state->crtc);
-	if (drm_WARN_ON_ONCE(dev, !crtc_state))
+	if (drm_WARN_ON_ONCE(dev, IS_ERR(crtc_state)))
 		goto err_drm_gem_end_shadow_fb_access;
 	sysfb_crtc_state = to_drm_sysfb_crtc_state(crtc_state);
 
-- 
2.51.0

Re: [PATCH next] drm/sysfb: Fix a NULL vs IS_ERR() bug

[Ville Syrjälä]

On Fri, Oct 17, 2025 at 07:03:33PM +0300, Dan Carpenter wrote:
> The drm_atomic_get_crtc_state() function never returns NULL, it returns
> error pointers.  Update the error checking to match.
> 
> Fixes: cb71de092553 ("drm/sysfb: Lookup blit function during atomic check")
> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
> ---
>  drivers/gpu/drm/sysfb/drm_sysfb_modeset.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
> index 8517c490e815..d2c4d8f3d4d0 100644
> --- a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
> +++ b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
> @@ -259,7 +259,7 @@ int drm_sysfb_plane_helper_begin_fb_access(struct drm_plane *plane,
>  	ret = -EINVAL;
>  
>  	crtc_state = drm_atomic_get_crtc_state(plane_state->state, plane_state->crtc);

Looking at where this gets called, that should be
drm_atomic_get_new_crtc_state(). Either that or the
code is more seriously borked.

I'll send out a patch that warns about invalid use of 
*_get_state()...

> -	if (drm_WARN_ON_ONCE(dev, !crtc_state))
> +	if (drm_WARN_ON_ONCE(dev, IS_ERR(crtc_state)))
>  		goto err_drm_gem_end_shadow_fb_access;
>  	sysfb_crtc_state = to_drm_sysfb_crtc_state(crtc_state);
>  
> -- 
> 2.51.0

-- 
Ville Syrjälä
Intel

Re: [PATCH next] drm/sysfb: Fix a NULL vs IS_ERR() bug

[Dan Carpenter]

On Fri, Oct 17, 2025 at 07:15:47PM +0300, Ville Syrjälä wrote:
> On Fri, Oct 17, 2025 at 07:03:33PM +0300, Dan Carpenter wrote:
> > The drm_atomic_get_crtc_state() function never returns NULL, it returns
> > error pointers.  Update the error checking to match.
> > 
> > Fixes: cb71de092553 ("drm/sysfb: Lookup blit function during atomic check")
> > Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
> > ---
> >  drivers/gpu/drm/sysfb/drm_sysfb_modeset.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
> > index 8517c490e815..d2c4d8f3d4d0 100644
> > --- a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
> > +++ b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
> > @@ -259,7 +259,7 @@ int drm_sysfb_plane_helper_begin_fb_access(struct drm_plane *plane,
> >  	ret = -EINVAL;
> >  
> >  	crtc_state = drm_atomic_get_crtc_state(plane_state->state, plane_state->crtc);
> 
> Looking at where this gets called, that should be
> drm_atomic_get_new_crtc_state(). Either that or the
> code is more seriously borked.
> 

I can't comment on that.  Let's drop this patch and instead (probably
I guess?) change it to drm_atomic_get_new_crtc_state().

regards,
dan carpenter

Re: [PATCH next] drm/sysfb: Fix a NULL vs IS_ERR() bug

[Thomas Zimmermann]

(cc dri-devel)

Am 19.10.25 um 17:06 schrieb Thomas Zimmermann:
> Hi
>
> Am 17.10.25 um 18:15 schrieb Ville Syrjälä:
>> On Fri, Oct 17, 2025 at 07:03:33PM +0300, Dan Carpenter wrote:
>>> The drm_atomic_get_crtc_state() function never returns NULL, it returns
>>> error pointers.  Update the error checking to match.
>>>
>>> Fixes: cb71de092553 ("drm/sysfb: Lookup blit function during atomic 
>>> check")
>>> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
>>> ---
>>>   drivers/gpu/drm/sysfb/drm_sysfb_modeset.c | 2 +-
>>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c 
>>> b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
>>> index 8517c490e815..d2c4d8f3d4d0 100644
>>> --- a/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
>>> +++ b/drivers/gpu/drm/sysfb/drm_sysfb_modeset.c
>>> @@ -259,7 +259,7 @@ int 
>>> drm_sysfb_plane_helper_begin_fb_access(struct drm_plane *plane,
>>>       ret = -EINVAL;
>>>         crtc_state = drm_atomic_get_crtc_state(plane_state->state, 
>>> plane_state->crtc);
>> Looking at where this gets called, that should be
>> drm_atomic_get_new_crtc_state(). Either that or the
>> code is more seriously borked.
>
> Yes that should be get_new_crtc_state. Thanks for spotting the bug.
>
>>
>> I'll send out a patch that warns about invalid use of
>> *_get_state()...
>
> If that is possible, I'd welcome it very much. State handling seems 
> like one of the dark corners of the DRM framework.
>
> Best regards
> Thomas
>
>>
>>> -    if (drm_WARN_ON_ONCE(dev, !crtc_state))
>>> +    if (drm_WARN_ON_ONCE(dev, IS_ERR(crtc_state)))
>>>           goto err_drm_gem_end_shadow_fb_access;
>>>       sysfb_crtc_state = to_drm_sysfb_crtc_state(crtc_state);
>>>   --
>>> 2.51.0
>

-- 
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Frankenstrasse 146, 90461 Nuernberg, Germany
GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
HRB 36809 (AG Nuernberg)